DroneFINE: Domain-Aware Parameter-Efficient Fine-Tuning of Vision-Language Detectors for Drone Images
2026-07-01 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors focus on improving object detection for drones, which is hard because current vision-language models (VLMs) are trained mainly on regular ground-level images, not bird's-eye views from drones. They found existing fine-tuning methods don't work well due to differences in image type and object size. To fix this, they designed DroneFINE, which includes a special adapter (HyperAdapter) that adjusts to drone image features and a guidance tool (SemanticGate) that helps ignore distracting background areas. Their tests show DroneFINE works much better than previous methods and almost as well as fully retraining the models, but with less training needed.
Unmanned Aerial VehiclesVision-Language ModelsParameter-Efficient Fine-TuningDomain GapBird's-Eye ViewForeground and BackgroundHyperAdapterSemanticGateObject DetectionDrone Imagery
Authors
Ke Wu, Yanan Zhang, Yingjie Gao, Wenhao Li, Chenyu Zhou, XinZhu Ma, Jiaxin Chen, Di Huang
Abstract
Object detection for Unmanned Aerial Vehicles (UAVs) working in open and dynamic environments is a highly challenging task. While Vision-Language Models (VLMs) have offered a powerful solution for universal object detection, adapting them to UAV scenarios remains non-trivial due to a substantial domain gap between VLM pre-training data and aerial imagery. The prevailing Parameter-Efficient Fine-Tuning (PEFT) methods prove ineffective in bridging this gap, as VLMs' "natural-scene, foreground-dominant" visual priors misalign with the "bird's-eye-view, background-dominant, small-object" characteristics of UAV data. To address this issue, we propose DroneFINE, a novel PEFT paradigm comprising two domain-aware complementary modules tailored for VLM-based drone image detectors. Specifically, a data-dependent, foreground-aware, and multi-path adaptation mechanism named HyperAdapter is designed, which overcomes the static structural constraints of PEFT. In addition, a background suppression algorithm named SemanticGate is developed. It is a text-conditioned guidance strategy that employs background vocabulary to actively guide the model in suppressing responses from irrelevant regions. Extensive experiments on VisDrone and UAVDT demonstrate that DroneFINE significantly outperforms existing PEFT methods and achieves performance comparable to full fine-tuning while substantially reducing the number of trainable parameters.